LED Grow Light
 

Last year's garden had a lot of problems, including an inadequate seedling grow light. Last year, my seedlings were really leggy and weren’t developing very well. I was forced to put them outside way too soon, and those conditions killed them quickly. Basically, I forgot to bring them in one too many nights (due to the cold) and forgot to water them one too many mornings. I wanted to remedy this and upgrade to LEDs, but still be cost effective. FYI, you can really learn a lot about indoor growing from watching the pot grower’s YouTube videos and reading some of their forums.
I ordered 13 – 10W LED full spectrum (400nm, 840nm) COBs from the old flea bay for $19.88. I ordered the 10W COBs because they had the lowest $/Watt at $0.11/Watt.

10w 20w 30w 50w 100w COB full spectrum led chip 400nm~840nm plant grow light #B | eBay

I also ordered 12 – 10W DC to DC constant current LED drivers for $13.32.

Practical 10W 12V- 24V DC LED Constant Current Driver Power 900mA High Power LED | eBay

In addition, I ordered Chinese-ium grade “thermal adhesive” from AliExpress for $0.58, including shipping. For $0.58 you can’t go wrong in a cheap project like this!

From Lowes, I got 3’x1.5”x1/8” Aluminum flat bar for $8.78 (ouch!) and aluminum 23 gauge 2’x6” sheet for $15.61 (double ouch!!).

Fortunately, I am a dumpster diver, so I had an old desktop computer supply (actually 3) laying around that I wanted to base this whole project around (since those are usually the biggest cost). Also, I had scavenging parts from a free rear projection TV I found on Craig’sList, which included two very beefy heat sinks.

For my LED grow light design, I wanted the aluminum bar to be the backbone and the main heat conduction path for the LEDs and heatsinks. I also wanted to be able to remove the reflector, if needed, and service the LEDs if they ever burned out. I cut the 3’ aluminum bar down to 18” (so I can make two lights if I want), drilled and tapped it, and laid out the LED COBs.

http://i78.photobucket.com/albums/j1..._Placement.jpg

I had a bunch of old automotive wire laying around from when I scrapped out my “race car” and soldered wires to the LED COBs and adhered them to the aluminum bar.

http://i78.photobucket.com/albums/j1...LEDs_Wired.jpg

http://i78.photobucket.com/albums/j1...asted_Dwon.jpg

Just a note about the “thermal adhesive.” I tested some adhesive the night before on two small pieces of aluminum and it did not break loose very easily at all, and it took prying with pliers to break the bond. I would not recommend bonding a heavy heat sink with just the thermal adhesive, but something small and lightweight like an LED COB, it is perfect!

I cut the aluminum reflector to be mounted on the bottom side of the aluminum bar and fit around the LED COBs.

http://i78.photobucket.com/albums/j1.../Reflector.jpg

The LED drivers are quite small. I put electrical tape on the back of them all, zip-tied them back to back, creating three legs, and wired them up to the power supply.

http://i78.photobucket.com/albums/j1...LED_Driver.jpg

http://i78.photobucket.com/albums/j1...s_Attached.jpg

Everything wired up and ready for a test run.

http://i78.photobucket.com/albums/j1...ete_Wiring.jpg

Blindingly bright! I was wearing cheap sunglasses and the color still messed with my eyes a bit.

http://i78.photobucket.com/albums/j1...n/IMG_2859.jpg

As a very poor comparison, here is what my old 2’ long 2 - T12 bulb florescent grow light setup looks like (only one was of the “grow” spectrum variety because they are so expensive):

http://i78.photobucket.com/albums/j1...n/IMG_2862.jpg

My old florescent setup would draw about 36W, while the new LED setup hovered around 75W. I still have to attach the heat sinks and reflector but I am quite satisfied with the results so far. Also, I am going to see if the heat sinks are adequate for passive cooling or if I need to attach some small CPU fans. So far, the total expenses for this project is $58.17, but priced out per unit it is actually $37.18. I believe I spent close to $30 for my 2’ florescent setup, including a single “grow” variety T12 bulb. When I build a second light setup, I will likely just use plated sheet steel for the reflector instead of aluminum. I originally wanted the reflector as part of the heat dissipation, but as I got into the build, I determined that was not going to be feasible. That should drop the unit cost well below $30. I will post up pictures of the finished product once it is complete.

A bunch of Youtube videos suggest that CFLs just close enough to the plants to be as bright and direct as possible but not close enough to cause heat damage seem to be the most cost effective option and they usually have the broad spectrum needed. I've heard some LEDs lack the proper spectrum and some come at a very high cost.

On average, how long do CFLs last....maybe 3 years of heavy use? A 60W CFL grow light (equivalent to 250W) costs about $11. From what I have seen, the lifespan of a CFL is a maximum of about 15,000 hours, or 1.7 years of continuous 24x7 running. The 10W LED COBs I got are "rated" (I know these COBs are not perfect and are grade B and lower) at 50,000 hours or 5.7 years of continuous 24x7 running. Even if each LED COB obtains 50% of its expected life span I am still doing better than the CFL.

Now lets talk about premature failure. What happens when the transformer in a CFL goes out early, are you going to fix it? I can have up to 3 of my LED COBs or their power drivers go out and still come out ahead on cost to replace. Also, what happens if you accidentally drop your CFL? You are immediately out $11, plus any cleanup. That happened to me last year and replacing a T12 bulb isn't an expense I want to have very often.

I am no pot grower, but I know those people know what they are doing far more than a typical "grow a few veggies indoors home gamer". They have some pretty impressive setups (as you can see on YouTube by searching for "LED grow lights"), and most of them are HPS or LED. They are ditching their HPS for LED because, according to them, they are far superior and can be tuned to the properly wavelengths for optimal growth.

Here is one of the first pot videos (there are TONS BTW) I clicked on:

https://www.youtube.com/watch?v=eS-48P4t9EA

This niche "LED grow light" market got supercharged a few years ago when Cree released the CXA line of Xlamp chip LED modules, then the CXB line. These modules pack a massive amount of light output into a very small package, rivaling HID light intensity and efficiency. As with many name brand products, they do the job well and ain't cheap. Other manufacturers are offering lots of "knockoff" products that attempt to rival the Cree modules at a discount. Cree has not sit idle, and has cross-licensed and invested in some of these rival companies to expand the industry as a whole. The Epistar/Epileds modules used in this build are a direct product of this strategic alliance. So much for the "Cheap Chinese junk" argument.

For growing seedlings, the light intensity should fall into a range mainly dictated by the plant being started. Spectrum is not as important in early stages of growth. The "purple/magenta" spectrum work 99% as good as the 3500K to 5000K "white" spectrum. Keeping the intensity where it needs to be for the seedling is WAAAY more important: not enough and you have spindly twigs, too much and you get bleached or scorched leaves. With these 10W chips, I imagine you could bleach and scorch seedlings pretty easily if you weren't careful.

I rigged up a single "warm white" 10W chip like this in my truck awhile ago to replace the burnt out dome light, and was amazed at just how much intensity actually comes out of it. I also figured out that the little regulator boards are dimmable. For a seedling shelf, it would be awesome to make a static mount for the lights and just dim or brighten them at will. Eliminating a chain or cable rig to raise and lower the lamp assemblies would take a whole lot of hassle out of the process.

[QUOTE=jeff5may;53601]The Epistar/Epileds modules used in this build are a direct product of this strategic alliance. So much for the "Cheap Chinese junk" argument.


Good to know! I was vaguely familiar with some of these goings-on but thanks for all of the detail. It is reassuring that my few $ aren't truly going to waste.

Again, good to know about the color spectrum. I have a suspicion that I will have too much light intensity, but I figured it is a lot easier to unplug LED COBs than it is to take the whole setup down to attach more on later.

Great idea! Really and truly, this grow light build was just a launching point for me to get back into electronics work. I am very much a "nuts and bolts, not volts" type of guy, so this project was more of an educational and experimental project than anything else. I hadn't picked up my soldering iron in years and yesterday was the first time I tested the current draw with my meter I purchased 3 years ago... I definitely want to expand this project, in the future, to have much easier usability, control, and functionality but still keep costs down. Eventually, I am going to use some version of this setup for an indoor aquaponics system.

The regulator boards you have look just the same as what I put in the truck. The typical circuits look like this:
http://www.circuitdiagramworld.com/u...12033003-4.jpg
The chip comes in 2 different packages:
http://datasheetcafe.databank.netdna...115-pinout.gif
To control the brightness, the DIM pin is used. For a single board, a resistance can be connected to pull the pin low. The chip has an internal 5VDC supply that runs the current sensing circuitry and the switching output. The DIM pin is tied to this 5V reference through a 200k ohm resistor. A 250K potentiometer can be connected between the DIM pin and ground to control the brightness.

To control multiple boards, you can feed them a common PWM signal to make them all behave the same. It would be easy to construct using a 555 timer ic and a few parts. I'm sure if you looked around on the web, another dirt cheap board could be found that could be wired straight to the DIM pin to supply a PWM signal.

Fascinating! I think I am going to see if I can hack one of the boards and add a pot on it.

As for updates, here is the finished product:

http://i78.photobucket.com/albums/j1...n/IMG_2869.jpg

http://i78.photobucket.com/albums/j1...n/IMG_2870.jpg

I am pretty satisfied with the end product. Will it be overkill for growing seedlings, yes, absolutely! Will it be overkill for some DWC hydroponic lettuce growing experiments, probably not too much....but that is what experimentation is for.

Nice how-to, and Quite a bit less expensive than a typical AeroGarden. On an AeroGarden, you raise the lamp as the seedlings grow in order to maintain light intensity relative to the seedlings leaf canopy.

I looked at the drivers last night. They do have the 4115 chip on them, and the DIM pin (pin 8) is not connected. I ordered a cheap 250k pot from Ebay, for a $1.50. Once I get it, I will experiment with the dimming functionality a bit.

Thanks, I hope it can help someone. Initially, my "dimming" will simply be only powering 1 LED COB at a time. If I need more I will just simply power more. As you can see what I am discussing with jeff5may, I may be able to dim the LED COBs with these exact driver boards with an external pot, which would eliminate the need to raise the grow light.

I don't see why it wouldn't be. I have ordered 2 things from AliExpress and have had 0 issues. Me personally, I tend to order more stuff from Ebay because their search engine is better at finding exactly what one needs without going through all of the dross.

Here is my initial setup. Only 1 LED COB hooked up to power and approx. 10" above the top of the seedling pots.

http://i78.photobucket.com/albums/j1...ial_Setup2.jpg

I was fortunate enough to plant some lettuce seeds for some other experimenting with the grow light, with no plans to plant them in the garden. In essence they were throw away seedlings, which ended up being a really good thing. The lettuce emerged a good 6 days before anything else and really indicated how much luminosity seedlings really need. Early on, I noticed the lettuce was getting leggy and powered a second, then a third LED COB on each consecutive day. You can see how leggy the lettuce was getting even with 3 10W LED COBs.

http://i78.photobucket.com/albums/j1...wth_3_1_17.jpg

At this point, I had all 3 LED COBs powered on the right side of the grow light. The heat sink above those three LED COBs was quite warm to the touch and the left heat sink was warm as well. At this point, I knew my other seedlings would be emerging very soon, so I sped up the changes and temporarily installed a CPU fan, powered a fourth LED COB, lowered the grow light to approx. 6" above the top of the pots, and put aluminum foil around the entire light.

http://i78.photobucket.com/albums/j1...inal_setup.jpg

http://i78.photobucket.com/albums/j1..._setup_top.jpg

It seemed to be helping but I went ahead and connected a fifth LED COB. Here is the growth last night.

http://i78.photobucket.com/albums/j1...wth_3_5_17.jpg

You can see on the foreground seedling stems are darkening, which is a good sign. Here you can see my other seedlings are starting to emerge.

http://i78.photobucket.com/albums/j1..._seedlings.jpg

I powered the sixth and final LED COB this morning, so the grow light is running full tilt now. I am hopeful that this will be sufficient and am really glad I went with my initial gut feeling of increasing the LED COBs from 4 to 6 on this setup. Also, I obviously had the wrong conclusion that this would be "Overkill," and have been proven otherwise.

Don't burn 'em. The first growth from the "seed leaves" is not especially light dependent, as the baby is feeding off the seed. In this phase, it is normal for the plants to stretch. It is also the time when the leaves are most sensitive to light. By the time the leaves show signs of stress, it's too late, and often the babies wilt and die.

Then again, you may not have enough output to do serious damage. I have zero practical experience growing with LED light sources. However, you have close to the same output as a 75 Watt HPS bulb, which will decimate seedlings up close. I have heard from others that the red+blue fixtures put out way more useful plant-growing light than they appear to.

So far, things are looking pretty good.

http://i78.photobucket.com/albums/j1...wth_3_7_17.jpg

As of right now, the grow light sits about 4" above the tallest seedling pots. A quick Internet search indicated one needs approx. 60W/sq ft of growth. I say approximately because that is for a Pot Plant (once again, they seem to be the only ones that do any scientific research in this stuff), and that may not apply as much to veggies. With the grow light about 4" above the top of the seedling pots, that puts a lighting footprint of 3.3 sq ft of illuminated area or 18 W/sq ft.

Comparing to images I took two days ago, it seems like I did have an inadequate amount of light. You can see there isn't much vertical growth in the seedling compared to 2 days ago, and the stem is darkening. I haven't noticed any indications of leaf burn or signs of stress but I will keep my eyes peeled.

http://i78.photobucket.com/albums/j1...wth_3_7_17.jpg

The real indicator that I am on the right track is when their second leaves start to emerge. I will be observant for signs of stress.

Any updates? How are the seedlings doing?

To be honest, not great. Nearly all of my seedlings still do not have their true leaves yet. Contrast that with my cucumbers and green beans I directly sowed in my garden only 20 days ago, they are shooting up like weeds.

Not all soils are created equally!
 

So, after a long delay I have finally gotten around to seeing what went wrong with my seedlings this year. I came up with 3 possible variables that caused all of my seedlings to fail:
1) Not enough light intensity from the grow light
2) Too cold in the garage and stunted growth
3) Poor soil

After replanting a bunch of basil seeds and pretty much getting the same result again, even though they were in direct sunlight, I figured I needed to test the soil.

I went to one of the big box hardware stores and got 3 other types of soil, allowing me to test 4 total soils:
1) Jiffy All Natural Seed Starter Mix
2) American Country Side Top Soil - My original soil that I used
3) Miracle Grow Nature's Care Potting Mix
4) Miracle Grow Seed Starting Potting Mix

For this test, I used the seeds that I had: Roma Tomato, Cherry Tomato, Jetstar Tomato, and Kale. Each pot had exactly 3 seeds in them, and they were all started at the same time, on 7 June 17.

Here is the experiment layout (from left to right in the pictures):
Pots #1) Jiffy All Natural, with seeds soaked for 1 hour.
Pots #2) American Country Side
Pots #3) Jiffy All Natural
Pots #4) MG Nature's Care
Pots #5) MG Seed Starting Mix

And here are the results as of today:
http://i78.photobucket.com/albums/j1...n/IMG_3099.jpg

A closer look at Pots 1 and Pots 2:
http://i78.photobucket.com/albums/j1...n/IMG_3100.jpg

A closer look at Pots 3, 4, and 5:
http://i78.photobucket.com/albums/j1...n/IMG_3101.jpg

From the results, it is quite clear that the soil I was using this past winter was garbage! Right now, the current winner is MG Nature's Care (the pots in the middle of picture #3), but I am going to wait a little longer before I can conclusively say it is the best seed starting medium. Also, I plan on running this exact same experiment under my grow light to see if I get similar results.

I used the mix the local commercial greenhouse uses to start all their fast growing annual stuff this year. It's like generic pro mix.

Thanks for the updates Nate & Jeff. At some point, the 20+ acres of farmland my wife and I have may benefit from this soil and lighting data. For that matter, the 10 bare root blue spruce seedlings I planted last summer may have benefit from the soil data, had I known it at the time...

Glad I could help. It is crazy how quickly technology has transformed. Even though I did this project earlier this year, these LED COBs are old technology. If I was to build this grow light again, I would go with these:

110v led cob grow light | eBay

And the latest picture as of last night:
http://i78.photobucket.com/albums/j1...n/IMG_3120.jpg

Any guesses as to which one was my original soil?

Anyways, right now I would say Jifffy All Natural seems to have faster growth but MG Nature's Care seems to have a better propagation rate. Ultimately, I think any growth medium is better than the cheap stuff.

Also, I did get this same (well nearly the same, a couple of different plants since I ran out of seeds for two) experiment setup under my grow light last night. As soon as they start sprouting (7-9 days from now) we will see how well the grow light does compared to actual sunlight.

The babies are coming along nicely. Your lighting set up is legit. Looks like you just need a few lava lamps and you're all set lol!

I started my grow lamp experiment on June 20th. This time I used Roma, Rio Grande, Brandywine, and Kale seeds (all I had left).

So, I go out of town Tuesday morning, leaving the wife back to take care of everything and come home, Friday night (June 30), to this:

http://i78.photobucket.com/albums/j1...n/IMG_3146.jpg

I believe this is pretty conclusive that the grow light does in fact work, but my seed starting soil I used this past winter is total and utter garbage (column 2 in the picture). One thing I will say is my grow light does seem under powered. I believe I should be able to mount 4 more 10W (9W actual) LED COBs on it to bring up the total output to around 100W (90W actual).

Also, this picture also shows that the Jiffy All Natural Seed Starting Mix (rows 1 and 3) is the best seed starting mix out of all of them. I forgot to get a picture of the plants outside in the sun, which indicates the same thing, but I will do that tomorrow morning.

I can't wait to get these plants in the garden as soon as they grow a little bit more and harden off!

Here are a couple of top down views. The first one is plants from under the growlight. The second are sun grown plants. Note, they were started at different times, so a comparison between growlight vs. sun light is not a good one.

For the soils, from left to right:
#1 - Jiffy All Natural w/ seeds soaked for 1 hour
#2 - American Country Top Soil (original)
#3 - Jiffy All Natural
#4 - MG Nature's Care
#5 - MG Seed Starting Mix

Growlight Plants:
http://i78.photobucket.com/albums/j1...n/IMG_3147.jpg

Sun Grown:
http://i78.photobucket.com/albums/j1...n/IMG_3149.jpg

Conclusions:
1) Jiffy All Natural is the best seed starting mix out of the 4 that I tried, followed by MG Seed Starting Mix.
2) In both experiments, I soaked one entire set of seeds(column #1 for both) for 1 hour and saw little to no difference in the final results. Longer seed soaking may be a future experiment.
3) A 50W (actual) LED COB grow light, running 19 hours a day, is just barely enough power to get a small set of seedlings started.

Good idea about the color spectrum,it will really help us to control the brightness.

Yeah, those pot heads really know their technology and have really pushed the grow-light market technology development! I think it would depend on a few different circumstances if I was to agree with "not using homemade."
First, if the crops you are growing are for the upcoming garden season and you cannot weather a seed sprout failure, yes I would agree.
If you are just playing around and can handle failures, I would disagree.
If you want to learn something new or test your skills, I would disagree.

Anyways, LED COB technology has continued to develop even after my initial project. Any new LED grow light projects I will use these:
https://www.ebay.com/itm/Full-Spectr...ydC01Ze8_OTyjA
They hook directly up to 110v AC, so you can literally plug them strait into the wall (obviously after you mount them to a heatsink).

LED Grow Light Version 1.1
 

Well I was one of those "lucky" feds who got a 35 day paid vacation and had plenty of free time to get around to upgrading my LED grow light. I added the 4 additional COBs and drivers then tidied everything up. I got the wires covered in heat shrink and the fan permanently mounted on a shroud. Also, I want to somehow cool the drivers with a little forced air because they got quite hot last year and I had a failed driver. I will probably get some cheap plastic tube from the hardware store and mount it directly above the fan and duct the airflow around the drivers.

Now the total output is approx 100W. While this might seem high I believe it will give me some flexibility on positioning the grow light that I previously didn't have.

https://i78.photobucket.com/albums/j...204_172002.jpg

http://i78.photobucket.com/albums/j1...204_171950.jpg


LED Grow Light Version 2.0 coming soon!

You can come here to find LED plant lights

Any updates Nate? I'm going to use some of these cheap COB Led for a diversion load from my solar. Since they are current driven devices, they should work fine for extra solar power.

Don't use photo bucket anymore. The blurry / watermark pics kill this great thread.

I have good luck with imgur for hosting. They seem to be the most reliable over time.